# Formwink — The form builder that learns your business. > Tell Formwink who your ideal customer is. Build forms that ask the right questions, hand sales a brief and a fit read on every inquiry, and show marketing which forms turn into pipeline. Source: https://formwink.ai/ The form builder for sales and marketing teams # Your best form hasn’t been built yet. Tell Formwink who your ideal customer is. Build forms that ask the right questions, hand sales a brief and a fit read on every inquiry, and show marketing which forms turn into pipeline. [Join the beta](/beta) [See how it learns](/how-it-learns) For small and mid-size businesses. No engineering needed. Free during the beta. Loomwork · an AI and automation studio Sample ✓ Who you want Growing teams with a repeatable workflow to automate. ? What your form asks “What would you like to take off your team’s plate?” Then follow-ups that fit the answer. ↻ What it learns Which inquiries your team marks as good fits, and what they had in common. Inside the product · 30-second tour ## See how Formwink understands your business and builds a smarter form. [Open full screen →](/tour/formwink-tour.mp4) Concept preview of the workspace with sample data. ## One form, four stages Sample data · Loomwork, an AI and automation studio Business Questions Fit Learning ### Start with your customer. Not a blank canvas. Loomwork’s website becomes an editable business profile. Everything suggested from the site is labeled, and nothing counts until you review it. Services Needs your review Customer onboarding, support triage, internal handoffs Typical buyers Needs your review Growing businesses with operations or customer-success teams Approach Needs your review Connects existing tools; adds AI where useful ### Ask questions with a reason. Each recommended question explains itself. Branches are approved paths, not improvisation. What would you like to take off your team’s plate? Required Why this question? Opens on the problem, not on a service list. The answer picks the branch. Which tools are involved? Required Why this question? Compared with the tools Loomwork has approved for delivery. How often does this happen? Required Why this question? Repeatability is what makes an automation worth building. Do you have a budget in mind? Optional Why this question? “Help scoping a first project” is a fine answer. Budget isn’t the qualification story. ### Recognize the right inquiries. Facts, reasons, and what is still unknown. Every scenario below is sample data. Maya Sam Devin Jonas Maya Chen · Harbor Analytics Sample data ✓ What they said Onboarding handoff across HubSpot → Notion → Slack, about 60 new customers a month. Predicted, then reviewed Strong service fit → Qualified Budget and access still to clarify. Service fit is not commercial qualification. Sam Okafor · Fieldstone Advisors Sample data ~ What they said “Too much manual work, but we don’t know where to start.” Predicted, then reviewed Needs clarification → Not reviewed An unclear request is not an unqualified customer. One useful follow-up. Devin Park · Park & Co Sample data ✕ What they said Wants a standalone consumer mobile app. Predicted, then reviewed Outside current services → Not qualified Kept for human review rather than discarded. Jonas Weber · Weber Logistics Sample data ~ What they said “Internal handoffs” that is really a dispatch-system rebuild. Predicted, then reviewed Strong service fit → Corrected: Not qualified The owner corrected the prediction with one tap. That correction is what the learning uses. ### Learn what turns into business. Your reviews and recorded outcomes feed one suggestion you can review, compare, and test. Changes need approval and can be reversed. 200 saw the form 50 submitted 15% qualified conversion (30 ÷ 200) ICP-informed suggestion · not tested “Not sure yet” doesn’t necessarily mean a poor fit. In this sample, 9 of the 50 submissions chose “Not sure yet.” The team reviewed 4 of those 9 as qualified after a call. ## Start with your customer. Not a blank canvas. Paste your website or describe your business. Formwink drafts a business profile and an ideal-customer summary. Website-derived suggestions are marked as suggestions until you review them, so the profile is yours, not a guess. Ideal customer · Loomwork Growing businesses with a repeatable onboarding, support, or handoff workflow, running on tools Loomwork already delivers with. Budget is a scoping question, not a gate. - Customer-provided · confirmed - Website suggestion · needs review - Unknown · allowed ## Ask questions with a reason. The first question asks what to take off the team’s plate. Approved branches follow: onboarding, support, or handoffs each get their own follow-ups, and “Not sure yet” gets a clarifying question instead of a dead end. Every recommendation says why. Which tools are involved? Why this question? Compared with the tools Loomwork has approved for delivery. Do you have a budget in mind? Why optional? “Help scoping a first project” is a fine answer. Budget isn’t the qualification story. Ordering: initial hypothesis · Profile v1 ## Recognize the right inquiries. Every inquiry gets a brief and a fit assessment against your services and capabilities: what matched, what didn’t, and what is still unknown. Service fit is kept separate from commercial qualification. You define a good lead. Formwink helps you recognize one. Nothing is rejected automatically. ✓ Strong service fit Onboarding workflow · known tools · recurring · budget still to scope ✕ Outside current services Standalone consumer app · kept for human review, not deleted ? Needs clarification “Too much manual work, not sure where to start” · one useful follow-up ## Learn what turns into business. Form behavior, your reviewed lead quality, and outcomes you record feed into recommendations with their evidence attached. Review the recommendation, compare versions side by side, and set up a test that measures qualified inquiries per visitor. Form behavior Reviewed quality Recorded outcomes ↓ “Not sure yet” doesn’t necessarily mean a poor fit. 4 of 9 unclear-scope inquiries were later reviewed qualified. A suggestion, not yet a result. Version A Not sure yet → free text → Version B Not sure yet → “most manual effort?” [See how it learns](/how-it-learns) ## Improve with evidence. Stay in control. An observation is not a result. Formwink keeps the two apart, shows when there isn’t enough data yet, and never changes a published form on its own. Every change is a version you approve and can reverse. ICP-informed suggestion Add a clarifying question for “Not sure yet.” Informed by your profile and 4 of 9 reviews, not yet a learned pattern. Tested improvement None yet. A test compares versions on the same metric with the same criteria before anything is called better. Insufficient data New forms show “Still learning” with the next useful action instead of a made-up recommendation. Simulated test setup: No experiment results yet. A suggestion is never shown as a result. Last 90 days · Form v1 · Profile v1 · Sample data. Email · Slack · Sheets · HubSpot · Webhooks ## The useful version, where your team already works. A readable brief with the fit assessment goes to your team. The respondent gets a plain confirmation. Internal assessments never reach prospects. [See all integrations](/integrations) #inquiries · Loomwork Preview formwink APP · 09:14 New inquiry: automate customer onboarding · Maya Chen, Harbor Analytics Harbor Analytics wants to connect HubSpot → Notion → Slack and reduce manual handoffs. About 60 new customers a month. Strong service fit · ✓ Onboarding · ✓ Known tools · ✓ Recurring · ? Budget and access Good fit Not a fit Not sure ## Templates with a goal attached [All templates →](/templates) [Studios & agencies Studio project inquiry Goal: Reviewed qualified inquiries per visitor](/templates/agency-client-inquiry) [Studios & agencies Website project brief Goal: Briefs complete enough to scope without a discovery call](/templates/website-project-brief) [Consultancies Fractional leadership inquiry Goal: Engagements with companies at the right stage](/templates/fractional-leadership-inquiry) ## Two plans. Both start with your ideal customer. ### Free $0 forever 1 published form · 100 submissions / month · Fit checks on every submission · 10 AI briefs / month Fit explanations and basic feedback [Join the beta](/beta) ### Pro $19 / month 10 published forms · 1,000 submissions / month · Fit checks on every submission · 100 AI briefs / month Conversion insights and reviewable form recommendations [Join the beta](/beta) Planned launch pricing. Formwink is in beta; paid plans are not yet available and terms may change. [Compare plans in detail →](/pricing) ## Questions people ask Short answers. Ask us anything else when you join the beta. ### Does Formwink change my form on its own? No. Recommendations wait for your review. A change becomes a new version only when you approve it, and any version can be rolled back as another version. ### What does “learning” actually use? Your ideal customer profile, form behavior on your own forms, your reviewed qualification labels, and outcomes you record or connect. Learning is specific to your workspace. ### Does the AI decide who is a good lead? It explains how an inquiry matches criteria you set and flags what is unknown. You confirm or correct it. No inquiry is rejected automatically. ### Do I need a CRM or Slack? No. Email delivery and manual outcome entry are enough to start. Slack, webhooks, and CRM connections are optional. ### Will my form get more submissions? That is not what we promise. Formwink helps you measure reviewed qualified inquiries per visitor and test changes against that. Results depend on your traffic, offer, and review habits. ### What happens to respondent data? It stays in your workspace. Internal fit assessments are never sent to respondents. We do not train shared models on your submissions. ## Give your forms something to learn from. Formwink is in beta. Everyone who applies gets an invite — fit only orders who we talk to first. [Join the beta](/beta) [Watch the 30-second tour](#tour) Reviewed · Separated · Labelled · Measured